Sunday Robotics’s new home robot, Memo, showed off impressive dexterity during a demonstration, such as making espresso, clearing glasses from a table, and loading a dishwasher, all tasks that traditional robots struggle with in messy, real-world settings.
This article from Wired.com tells that Memo moves on wheels and adjusts its height via a central column, rather than trying to ambulate on legs. That design choice keeps things simpler and more stable for a home environment. More interesting is how Sunday trains Memo: instead of laborious teleoperation, remote workers wear gloves that mimic Memo’s hands as they perform chores. The sensor data then teaches the robot how to replicate those motions with its own arms.
The article emphasizes the gap between most robots, designed for repetitive, well-structured industrial environments, and what you need for a home: adaptability, perception of diverse objects, flexible grasping, and recovery from unplanned situations. Memo’s ability to pick up two fragile wine glasses in one hand illustrates the leap in dexterity achieved via that glove-driven data collection method.
Sunday plans to begin beta testing Memo next year, targeting early adopters comfortable with some rough edges in exchange for a truly home-assistant robot. Full-stack integration, such as hardware plus data collection plus AI models, is emerging as the viable path to credible home robots, not just flashy demos.
The era when a machine could reliably handle everyday household chores is moving from lab demos into real-home experiments. This could reshape both the smart-home market and expectations around what “help at home” looks like.